Update: my BCI demo at a Des Moines VA clinic flopped until a nurse said one thing
I volunteered to demo a basic noninvasive BCI headset at a VA clinic in Des Moines last week, 8 channels on a cap, thinking the vets would be wowed by cursor control. A nurse named Dana pulled me aside and asked why I was showing off instead of solving anything, and it hit me that all my demos skip the boring part of actually fitting the hardware for each person. I spent the rest of the day just letting four guys try the calibration themselves, and 2 of them got steady signals in under 10 minutes once I shut up. Has anyone else found that the real bottleneck with noninvasive headsets is the setup, not the decoding?